Xiuxiu Hao, Baoqi Zhang, Lei Zhang, Yang Zhang
The CIL index was strongly associated with concurrent CT-FFR status within this cohort. Because the index and endpoint were derived from the same CCTA examination, and no external validation, invasive physiological reference, or longitudinal outcomes were available, the results represent within-examination association and internal discrimination rather than independent diagnostic accuracy, prognosis, or established clinical utility.
BACKGROUND: Coronary computed tomography angiography (CCTA) depicts coronary anatomy, plaque composition, and pericoronary inflammation, whereas CT-derived fractional flow reserve (CT-FFR) provides a computational functional estimate. We examined whether a prespecified plaque component-inflammation load (CIL) index was associated with concurrent CT-FFR-defined functional limitation.
METHODS: This single-center retrospective cross-sectional study included 334 patients (336 vessels or lesions) undergoing clinically indicated CCTA with quantitative plaque analysis, fat attenuation index (FAI) measurement, and CT-FFR computation. The CIL index was the equal-weight mean of standardized lipid component volume, fibrofatty component volume, lipid-to-fibrous ratio, and mean FAI. The primary patient-level endpoint was CT-FFR ≤0.80 in the lesion with the lowest CT-FFR value. Multivariable association, repeated stratified five-fold cross-validation, calibration, bootstrap optimism correction, and prespecified sensitivity analyses were performed.
RESULTS: Concurrent CT-FFR ≤0.80 was present in 119 patients (35.6%). A 1-standard-deviation higher CIL index was associated with higher odds of CT-FFR ≤0.80 (adjusted odds ratio 10.60, 95% confidence interval 5.09-22.10; P < 0.001). The clinical-plus-anatomy-plus-CIL model had an internally cross-validated AUC of 0.940, PR-AUC of 0.891, and Brier score of 0.090; its optimism-corrected AUC was 0.945. Sensitivity analyses were directionally consistent. Cancer-related subgroup and treatment-interaction estimates were exploratory and imprecise.
CONCLUSION: The CIL index was strongly associated with concurrent CT-FFR status within this cohort. Because the index and endpoint were derived from the same CCTA examination, and no external validation, invasive physiological reference, or longitudinal outcomes were available, the results represent within-examination association and internal discrimination rather than independent diagnostic accuracy, prognosis, or established clinical utility.